Night Variation in Fibroblast Growth Factor 21 Levels in Young

ORIGINAL ARTICLE Lack of Day/Night variation in ﬁbroblast growth factor 21 levels in young healthy men

J-P Foo1, KN Aronis1,2,3, JP Chamberland1,2 and CS Mantzoros1,2

BACKGROUND: Fibroblast growth factor (FGF) 21 is an endocrine factor with an emerging role as a metabolic regulator. We previously reported the presence of a significant day/night variation of FGF-21 in energy-replete, healthy female subjects. However the day/night patterns of secretion in male subjects remain to be fully elucidated. To elucidate day/night pattern of FGF-21 levels in male subjects in the energy-replete state, its relationship to FFA and to investigate whether a sexual dimorphism exists in FGF-21 physiology. METHODS: Eight healthy lean male subjects were studied for up to 5 days while on an isocaloric diet. Blood samples were obtained for measurement of FGF-21 and free fatty acids (FFA) hourly from 0800 AM on day 4 till 0800 AM on day 5. RESULTS: FGF-21 did not exhibit any statistically significant day/night variation pattern of circulating FGF-21 levels during the isocaloric fed state in male subjects. FGF-21 levels in male subjects are closely cross-correlated with FFA levels, similar to female subjects. CONCLUSIONS: A sexual dimorphism exists in FGF-21 physiology; that as opposed to female subjects, no significant day/night variation exists in FGF-21 rhythm in male subjects in the energy-replete state. Circulating pattern of FGF-21, similar to the female subjects, was highly cross-correlated to the FFA levels in the male subjects, signifying that the sexual dimorphism in FGF-21 physiology may be related to the differing lipid metabolism in both the genders. International Journal of Obesity (2015) 39, 945–948; doi:10.1038/ijo.2014.215

INTRODUCTION particular its biological rhythm is unclear and inconsistent. A study Fibroblast growth factor (FGF) 21 is an endocrine factor with an reported the absence of any diurnal variation in FGF-21 in healthy 7 emerging role as a metabolic regulator.1 In mice, FGF-21 acts as a subjects, while another study reported the presence of a 9 potent regulator of glucose and lipid metabolism. In particular, circadian rhythm. These studies, however, had not examined FGF-21 levels are induced by fasting, via increased hepatic whether a sexual dimorphism exists in the biological rhythm of expression of peroxisome proliferator-activated receptor gamma FGF-21, and had reported the physiology of FGF-21 in both coactivator protein-1alpha,2 with corresponding increases in fatty genders simultaneously. We previously reported, via a robust acid oxidation and gluconeogenesis, highlighting the role of study design with high temporal resolution, the presence of a fi FGF-21 in the regulation of carbohydrate and fatty acid signi cant day/night variation of FGF-21 in energy-replete, metabolism during energy homeostasis.3 Administration of healthy, lean female subjects, and the close relationship of the process of lipolysis and circulating levels of FGF-21 in these female FGF-21 in diet-induced obese mice on the other hand reverses subjects.10 Given that significant sexual dimorphism in lipid the hepatic steatosis, reduces plasma glucose and triglycerides to – metabolism exists in humans,11 14 we hypothesize that the day/ near normal levels and improves insulin sensitivity,4 providing night variation pattern of FGF-21, which is closely related to lipid evidence of the therapeutic potential of FGF-21 in regulating metabolism, may be different in male subjects. glucose and lipid metabolism in rodents. In this study, we aim to elucidate, via a well-standardized study In humans, FGF-21 has been found to correlate positively with 10 fi design similar to the female study, the biological rhythm of adiposity, fasting insulin and triglycerides and is signi cantly FGF-21 in the energy-replete state, and to examine the relation- higher in obese than in lean subjects, raising the possibility of a 5,6 ship between FGF-21 and lipolysis in healthy male subjects, and FGF-21-resistant state in obese subjects. A study in humans whether a sexual dimorphism exists in FGF-21 physiology. reported a significant increase in FGF-21 levels during prolonged fasting, a process which increases free fatty acids (FFA), and which – consequently activates peroxisome proliferator activator recep- METHODS tor-α.7 In vivo, lipid infusion, likely via peroxisome proliferator– Eight healthy, non-diabetic, lean men (mean age 21.8 ± 2.3 years; mean activator receptor-α activation, induces an increase of circulating − 2 body mass index 21.2 ± 1.6 kg m ) were admitted to the Beth Israel FGF-21, further attesting a direct and significant interaction 8 Deaconess Medical Center General Clinical Research Center and studied in between lipid metabolism and FGF-21 regulation in humans. the isocaloric fed state for a total of 5 days. Similar to the female study,10 Despite the role of FGF-21 in metabolic regulation and energy these male subjects were screened via a detailed historic and systemic homeostasis, the fundamental physiology of FGF-21 in humans, in review and by baseline laboratory analysis and chronic conditions

1Division of Endocrinology, Diabetes & Metabolism, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA; 2Section of Endocrinology, Boston VA Healthcare System, Harvard Medical School, Boston, MA, USA and 3Department of Medicine, Boston Medical Center, Boston University, Boston, MA, USA. Correspondence: Professor CS Mantzoros, Harvard Medical School and Harvard School of Public Health, VA Boston Healthcare System, 150 S. Huntington Avenue, Boston, MA 02130, USA. E-mail: [email protected] Received 20 May 2014; revised 23 September 2014; accepted 1 October 2014; accepted article preview online 26 December 2014; advance online publication, 20 January 2015 FGF-21 physiology in young healthy men J-P Foo et al 946 including infectious diseases, renal or hepatic failure, Type 1 or Type 2 physically being in the subject's room. On day 4, blood was drawn diabetes mellitus, cancer, lymphoma, hypogonadism, malabsorption or through an indwelling intravenous catheter, every 15 min from 0800 AM till malnourishment, hypo or hyperthyroidism, hypercortisolism, alcoholism or 0800AM on day 5, and then pooled every hour to meet the assays’ sample drug abuse or subjects who have been on medications known to affect the volume requirements. hormones to be measured in this study were ruled out prior to the The study protocols were approved by the Institutional Review Board of recruitment. All the subjects had normal liver and renal function. There Beth Israel Deaconess Medical Center, and written informed consent was were no significant differences between fasting glucose (P40.05) or body obtained from all subjects. mass index (P40.05) between the male and female subjects. Efforts were also made to standardize their activities including their sleeping conditions Assays during the study period. All subjects were acclimatized to a research bed Serum FGF-21 was measured by enzyme-linked immunosorbent assay for 2 days prior to the frequent sampling day. The subjects were admitted − 1 on day 1 at 9 PM the night before the commencement of the study on day (R&D Systems, Minneapolis, MN, USA) with a sensitivity of 4.67 pg ml , fi – fi 2. The study concluded after the 3-day intervention and ended on Day 5 intra-assay coef cient of variability of 2.9 3.9 % and inter-assay coef cient – ’ when they were discharged after the last blood taking at 8 AM. Subjects of variability of 5.2 10.9%, in accordance with the manufacturer s − were given a standardized isocaloric diet with breakfast at 0800 AM, lunch instructions. All serum samples were stored at 80 °C as appropriate until at 0100 PM, dinner at 0600 PM and a snack at 1000 PM daily. Caloric intake analysis. All samples were analyzed in duplicate. FFAs levels were was distributed with 20% of calories from breakfast, 35% from lunch, 35% measured by an enzymatic colorimetric assays (Wako Diagnostic, Mountain from dinner and 10% from the evening snack. Subjects were allowed ad View, CA, USA). libitum physical activities that were not different from their usual activities in the outpatient settings. All light/dark intervals and blood sampling Statistical analysis schedule were standardized. All subjects were equally exposed to light for Statistical analysis was performed using Stata version 12 (Stata Corp. 16 h (7 am to 11 pm) and were in the dark for 8 h (11 pm to 7 am), so as to College Station, TX, USA) Descriptive statistics are presented as means ensure similar exposure to light/dark duration. To avoid disturbing the ± s.d. Normality of the variables was evaluated using the Shapiro–Wilkes subjects at night, research nurses performed blood draws without test. Variables that were not normally distributed were normalized using the appropriate, for each time, transformation. Four parameter, sine, ordinal least squares, regression analysis on the data from all subjects was 400 Adj R^2 = 0.2764 performed to estimate for potential oscillations on FGF-21 and FFA circulating levels. AUC of FGF-21 and FFA were estimated using the trapezoid method. Comparisons of FGF-21 and FFA mean and AUC levels ’ 300 between male and female was performed with Student s t-test. Cross- correlation analysis between FGF-21 and FFA circulating levels was performed, at the level of each individual, using the CORRELATION 200 algorithms of the Pulse XP software accordingly (UVA Pulse Analysis Software, Charlottesville, VA, USA). All P-values are two-tailed and the alpha criterion was set to 0.05. 100 FGF-21 (pg/mL) RESULTS

0 Energy-replete fed state Our four-parameter, sine, ordinal least squares regression analysis did not reveal any statistically significant day/night variation pattern -100 of circulating FGF-21 levels during the isocaloric fed state in these 04812162024 male subjects. The estimated oscillation of FGF-21 levels had a Time (hours from 8AM) period of 93.90 ± 281.96 h (P = 0.74), amplitude of 262.97 ± 1493.35 pg ml − 1 (P = 0.86) and mean hormonal levels of 311.41± 1796.56 pg ml − 1 (P = 0.84). The statistically non-significant amplitude and 1.2 Adj R^2 = 0.0360 period derived from our model suggest that circulating FGF-21 levels do not have any day/night variation pattern (Figure 1a). The 1.0 adjusted coefficient of determination of our final model was 27.64% (Po0.01), meaning that our model adequately described the data. 0.8 Application of our four-parameter sine model on FFA levels yielded statistically significant parameters: amplitude of 0.05 ± 0.02 mEq L − 1, 0.6 period of 28 ± 6.6 h and mean levels of 0.32 ± 0.1 mEq L − 1 (all P-values o0.01). The adjusted coefficient of determination, 0.4 however, was 3.6%, P = 0.02 (Figure 1b), implying that the FFA times FFA (mEq/L) 0.2 series is only poorly described by a periodic function and thus we could conclude that FFA do not exhibit any statistically significant 0.0 day/night variation pattern in these male subjects. In contrast, we previously reported via similar four-parameter, -0.2 sine, ordinal least squares regression analysis, the presence of a 0 4 8 12162024 significant day/night variation pattern of circulating FGF-21 levels Time (hours from 8AM) during the isocaloric fed state in young healthy, lean female − subjects. The estimated oscillation of FGF-21 levels had a period of Figure 1. (a) Day/night variation of mean FGF-21 levels (pg ml 1)in o − 1 2 27.86 ± 2.76 h (P 0.001), amplitude of 60.988 ± 5.866 pg ml the fed state. Adjusted coefficient of determination (R ) is displayed o − 1 fi (P 0.001), and mean hormonal levels of 98.75 ± 6.67 pg ml at the top center. Solid line represents 95% con dence interval, (Po0.001), with an adjusted coefficient of determination of interrupted line represents 95% prediction interval (n = 8); (b) day/ − o fi night variation of mean FFA (mEq L 1) levels in the fed state. 48.28% (P 0.001). The statistically signi cant amplitude and Adjusted coefficient of determination (R2) is displayed at the top period suggest that a significant proportion of circulating FGF-21 center. Solid line represents 95% confidence interval, interrupted levels variability could be explained by an underlying day/night line represents 95% prediction interval (n = 8). variation pattern in these female subjects. FFA 24-h pattern,

International Journal of Obesity (2015) 945 – 948 © 2015 Macmillan Publishers Limited FGF-21 physiology in young healthy men J-P Foo et al 947 similar to FGF-21, exhibited a significant pattern of day/night been conflicting, with one study demonstrating the absence of variation in the fed state with ordinal least squares regression any diurnal variation in healthy subjects,7 while other studies analysis in these female subjects.10 reporting its presence.9,22 Galman et al.7 reported the absence of diurnal variation of FGF-21 in five healthy subjects. However, the 9 Comparison of FGF-21 and FFA levels between Men and Women gender of these subjects was not reported. Yu et al., in contrast, reported the presence of a circadian pattern of FGF-21 in the fed There was no statistically significant difference in FGF-21 24 h AUC state. However, similarly, the gender of these subjects was not and mean levels between male and female subjects in the reported. Andersen et al.22 reported the presence of a circadian isocaloric fed state (P = 0.35 and 0.39 accordingly). Similarly, FFA rhythm of FGF-21 in healthy female subjects, but male subjects 24 h AUC and mean levels were similar between male and female were not studied. The inconsistent data with FGF-21 biological (P = 0.76 and 0.69 accordingly).10 characteristics hampers a clear understanding of its biological role in humans. Given the substantial interest in FGF-21 as a Relationship of FGF-21 with FFA therapeutic target in diabetes,23–25 the lack of knowledge on its Cross-correlation analysis demonstrated that the 24-h circulating periodicity of secretion has important ramifications on how pattern of FGF-21 was closely associated to the FFA 24-h pattern. various clinical studies on FGF-21, each utilizing different sampling Although the subjects were in the isocaloric fed state, FGF-21 and time and conditions can be interpreted. Furthermore, the FFA levels were strongly cross-correlated in seven out of eight circadian periodicity of metabolites in humans may play an male subjects, exhibiting cross-correlations with values ranging important role in the pathogenesis of metabolic disorder. Notably, from 0.31–0.59 and observed with 2–8 h lag. Similarly, we disrupted circadian rhythmicity of metabolic hormones has been previously reported that four out of six female subjects exhibited implicated in the development of obesity and metabolic significant positive cross-correlation ranging from 0.35–0.81 disorders.26–28 Knowledge of the circadian rhythmicity of an between FGF-21 and FFA levels at 2–6 h lag in the isocalcoric fed endocrine factor such as FGF-21 is imperative to understanding its state.10 physiological role in humans. We previously reported the presence of a significant day/night variation pattern of female subjects in the fed state.10 The current DISCUSSION study, done under similar strictly standardized condition as the We demonstrated for the first time via a well-standardized study female study, demonstrated the absence of such day/night design, that a sexual dimorphism exists in FGF-21 physiology; that variation in male subjects, hereby illustrating a sexual dimorphism as opposed to female subjects, no significant day/night variation in FGF-21 physiology. This sexual differences may very well be the in FGF-21 biological rhythm exists in male subjects in the energy- key point that underlies many of the conflicting data surrounding replete state. We further showed that the circulating pattern of its biological properties. Such an observation in sexual dimorph- FGF-21, similar to the female subjects, was closely cross-correlated ism of FGF-21 is not entirely unsurprising, given that significant to the FFA levels in the male subjects, suggesting that the sexual sexual dimorphism exist in lipids metabolism. Furthermore, it is dimorphism in FGF-21 physiology may be highly related to the now clear that FGF-21 is highly related to lipid metabolism in differing lipids metabolism in both genders. humans. This observation of a sexual differences in FGF-21 It is well reported that significant differences exists in lipid physiology herald a step forward in the understanding of the metabolism between both the genders. For one, it is a well-known physiology of FGF-21 and provides crucial insight on how future observation that women generally have a higher amount of body studies on FGF-21 and its therapeutic implications should be fat than men. A study found that women have a greater increase interpreted. in peripheral lipolysis in response to catecholamines compared Our study has several strengths. The male subjects and study with men, and that women are more sensitive to the lipolytic conditions were as highly standardized as the female study, with action of catecholamines,15 the level of which is increased in early all subjects receiving similar isocaloric diet, activity level, with starvation and energy deprivation.16,17 Nocturnal FFAs are well uniform exposure to light/dark interval, thus allowing us to make described to increase in the early hours during period of relative robust comparisons with the female study. Blood sampling was energy deprivation.18–20 In the transition from fed to fasting state, frequent, resulting in a high temporal resolution in our time series, insulin level falls, leading to release of FFA from adipose tissue to therefore enabling us to conclude on the presence of a day/night be utilized by the muscle and liver for energy production or to be pattern of FGF-21 secretion with great confidence. The FGF-21 stored as triglyceride. The rate of lipolytic release during fasting assay utilized in the study has been thoroughly evaluated and all and starvation exceeds whole-body energy requirements in a the assays were performed in duplicate by operators blinded to sexually dimorphic manner, with women having significantly the study hypothesis, eliminating any measurement bias. The higher release rates than men.21 In this study, although the 24 h study was adequately powered to detect a day night variation AUC and mean levels of FFA in men were similar to women, pattern. Prior studies of smaller sample sizes and the same significant differences can be observed in FFA secretion pattern in temporal resolution with the present study have been shown to both genders. In the early mornings, where the body is in a be adequately powered to demonstrate a statistically and clinically relative state of energy deprivation, the FFA levels were observed significant day/night variation in FGF-21 levels.10 The strict to spike to a higher level, giving rise to a more pronounced day/ standardization of study conditions enhances the study quality night variation in FFA secretion over 24 h in women.10 In men, by eliminating the uncontrolled confounding. although an increased in FFA levels were observed in the early Some limitations have to be addressed. Our study is confined to morning, this increase was less marked, resulting in an absence of lean healthy male subjects, and it is unknown if similar findings significant day/night variation pattern in the male subjects. And can be generalized to obese or diabetic subjects. The cross- given that FFA levels and FGF-21 levels were highly cross- correlationship of FGF-21 and FFA provides hypothesis on the correlated in both men and women, these results suggest that the interaction between lipolysis and FGF-21 physiology in humans. absence of significant day/night variation pattern of FGF-21 However, causality remains to be proven. observed in men in this study may be highly related to the In summary, we found for the first time, via a highly differences in lipolysis response in early starvation in both standardized study design, that a sexual dimorphism exists in genders. FGF-21 physiology; that young healthy men do not exhibit The absence of a day/night variation of FGF-21 levels in men as significant day/night variation pattern of FGF-21 levels as opposed opposed to women is a novel observation. Previous studies have to women. We further demonstrated, similar to the findings in

© 2015 Macmillan Publishers Limited International Journal of Obesity (2015) 945 – 948 FGF-21 physiology in young healthy men J-P Foo et al 948 women, that FGF-21 is highly cross-correlated to lipolysis in these 10 Foo JP, Aronis KN, Chamberland JP, Paruthi J, Moon HS, Mantzoros CS. Fibroblast men, providing support that the well-described sexual differences growth factor 21 levels in young healthy females display day and night in lipid metabolism may be the key factor in effecting the sexual variations and are increased in response to short-term energy deprivation dimorphism in FGF-21 physiology described herein. These novel through a leptin-independent pathway. Diabetes Care 2013; 36: 935–942. 11 Wang X, Magkos F, Mittendorfer B. Sex differences in lipid and lipoprotein observations provide critical insights on FGF-21 physiology in 96 humans. metabolism: it's not just about sex hormones. J Clin Endocrinol Metab 2011; : 885–893. 12 Jensen MD, Johnson CM. Contribution of leg and splanchnic free fatty acid (FFA) fl 45 CONFLICT OF INTEREST kinetics to postabsorptive FFA ux in men and women. Metab Clin Exp 1996; : 662–666. The authors declare no conflict of intrest. 13 Jensen MD, Cryer PE, Johnson CM, Murray MJ. Effects of epinephrine on regional free fatty acid and energy metabolism in men and women. Am J Physiol 1996; 270:E259–E264. ACKNOWLEDGEMENTS 14 Mittendorfer B, Horowitz JF, Klein S. Gender differences in lipid and glucose This study was supported by the National Institutes of Health National Center for kinetics during short-term fasting. Am J Physiol Endocrinol Metab 2001; 281: – Research Resources grant M01-RR-01032 (Harvard Clinical and Translational Science E1333 E1339. Center). The content is solely the responsibility of the authors and does not 15 Horton TJ, Dow S, Armstrong M, Donahoo WT. Greater systemic lipolysis in necessarily represent the official views of the National Center for Research Resources women compared with men during moderate-dose infusion of epinephrine or the National Institutes of Health. Clinical trial registration: NCT00140231. and/or norepinephrine. J Appl Physiol 2009; 107:200–210. 16 Zauner C, Schneeweiss B, Kranz A, Madl C, Ratheiser K, Kramer L et al. Resting energy expenditure in short-term starvation is increased as a result of an increase AUTHOR CONTRIBUTIONS in serum norepinephrine. Am J Clin Nutr 2000; 71:1511–1515. 17 Chan JL, Mietus JE, Raciti PM, Goldberger AL, Mantzoros CS. Short-term JPF wrote manuscript, researched data, performed lab work; KNA analyzed fasting-induced autonomic activation and changes in catecholamine levels are the data, wrote manuscript; JPC peformed lab work; CSM is the principal not mediated by changes in leptin levels in healthy humans. Clin Endocrinol (Oxf) investigator, designed and supervised the study and reviewed and edited the 2007; 66:49–57. manuscript. 18 Schlierf G, Dorow E. Diurnal patterns of triglycerides, free fatty acids, blood sugar, and insulin during carbohydrate-induction in man and their modification by nocturnal suppression of lipolysis. J Clin Invest 1973; 52:732–740. REFERENCES 19 Boyle PJ, Avogaro A, Smith L, Bier DM, Pappu AS, Illingworth DR et al. Role of GH 1 Kharitonenkov A, Shiyanova TL, Koester A, Ford AM, Micanovic R, Galbreath EJ in regulating nocturnal rates of lipolysis and plasma mevalonate levels in normal 263 – et al. FGF-21 as a novel metabolic regulator. J Clin Invest 2005; 115: 1627–1635. and diabetic humans. Am J Physiol 1992; : E168 E172. 2 Kim KH, Kim SH, Min YK, Yang HM, Lee JB, Lee MS. Acute exercise induces FGF21 20 Hagstrom-Toft E, Bolinder J, Ungerstedt U, Arner P. A circadian rhythm in lipid 40 – expression in mice and in healthy humans. PloS One 2013; 8: e63517. mobilization which is altered in IDDM. Diabetologia 1997; : 1070 1078. 3 Potthoff MJ, Inagaki T, Satapati S, Ding X, He T, Goetz R et al. FGF21 21 Koutsari C, Basu R, Rizza RA, Nair KS, Khosla S, Jensen MD. Nonoxidative free fatty induces PGC-1alpha and regulates carbohydrate and fatty acid metabolism acid disposal is greater in young women than men. J Clin Endocrinol Metab 2011; 96 – during the adaptive starvation response. Proc Natl Acad Sci USA 2009; 106: :541 547. 10853–10858. 22 Andersen B, Beck-Nielsen H, Hojlund K. Plasma FGF21 displays a circadian rhythm 75 4 Xu J, Lloyd DJ, Hale C, Stanislaus S, Chen M, Sivits G et al. Fibroblast growth factor during a 72-h fast in healthy female volunteers. Clin Endocrinol (Oxf) 2011; : – 21 reverses hepatic steatosis, increases energy expenditure, and improves insulin 514 519. sensitivity in diet-induced obese mice. Diabetes 2009; 58: 250–259. 23 Huang J, Ishino T, Chen G, Rolzin P, Osothprarop TF, Retting K et al. Development 5 Zhang X, Yeung DC, Karpisek M, Stejskal D, Zhou ZG, Liu F et al. Serum FGF21 of a novel long-acting antidiabetic FGF21 mimetic by targeted conjugation to a 346 – levels are increased in obesity and are independently associated with the scaffold antibody. J Pharmacol Exp Therapeut 2013; :270 280. metabolic syndrome in humans. Diabetes 2008; 57: 1246–1253. 24 Smith R, Duguay A, Weiszmann J, Stanislaus S, Belouski E, Cai L et al. A novel 27 – 6 Fisher FM, Chui PC, Antonellis PJ, Bina HA, Kharitonenkov A, Flier JS et al. Obesity approach to improve the function of FGF21. BioDrugs 2013; :159 166. is a fibroblast growth factor 21 (FGF21)-resistant state. Diabetes 2010; 59: 25 Hecht R, Li YS, Sun J, Belouski E, Hall M, Hager T et al. Rationale-based engineering 2781–2789. of a potent long-acting FGF21 analog for the treatment of Type 2 diabetes. PloS 7 7 Galman C, Lundasen T, Kharitonenkov A, Bina HA, Eriksson M, Hafstrom I et al. One 2012; : e49345. The circulating metabolic regulator FGF21 is induced by prolonged fasting and 26 Korkmaz A, Topal T, Tan DX, Reiter RJ. Role of melatonin in metabolic regulation. 10 – PPARalpha activation in man. Cell Metab 2008; 8: 169–174. Rev Endocr Metab Disord 2009; :261 270. 8 Mai K, Andres J, Biedasek K, Weicht J, Bobbert T, Sabath M et al. Free fatty acids 27 Gangwisch JE, Feskanich D, Malaspina D, Shen S, Forman JP. Sleep duration link metabolism and regulation of the insulin-sensitizing fibroblast growth factor- and risk for hypertension in women: results from the nurses' health study. 26 – 21. Diabetes 2009; 58: 1532–1538. Am J Hypertens 2013; : 903 911. 9 Yu H, Xia F, Lam KS, Wang Y, Bao Y, Zhang J et al. Circadian rhythm of circulating 28 Scheer FA, Hilton MF, Mantzoros CS, Shea SA. Adverse metabolic and fibroblast growth factor 21 is related to diurnal changes in fatty acids in humans. cardiovascular consequences of circadian misalignment. Proc Natl Acad Sci USA 106 – Clin Chem 57: 691–700. 2009; :4453 4458.